23andMe's current status and the main indicator of Alzheimer's risk

Where 23andMe stands in 2026

23andMe filed for bankruptcy in March 2025. On 27 June 2025 a federal bankruptcy judge approved the sale of substantially all of its assets, including the Personal Genome Service and the research business, to TTAM Research Institute, a non-profit founded by 23andMe's co-founder Anne Wojcicki, for $305 million. The purchase completed on 14 July 2025 and the service continues to operate under TTAM's ownership. Twenty-seven state attorneys general had sued to block the sale; the judge called selling genetic data "a scary proposition" but found no law prohibiting it. TTAM pledged to keep customers' right to delete their data and to strengthen privacy policies.

The company's own figure is that about 80 percent of its more than 15 million customers consented to research use of their data. In 2023 a credential-stuffing attack exposed data on about 6.9 million users, leading to a $30 million US class settlement and a £2.31 million fine from the UK Information Commissioner in June 2025.

If you hold a 23andMe account, the practical steps are to turn on two-factor authentication, review your research consent, and decide whether you want your data to remain on the service at all. Deletion is available from account settings.

What the Alzheimer's report actually tests

The "Late-Onset Alzheimer's Disease" Genetic Health Risk report was among the ten reports the FDA authorised on 6 April 2017 through its de novo pathway. It is indicated for reporting the ε4 variant of the APOE gene, and nothing else. In the company's words, the report does not identify or report on the ε2 and ε3 variants, and having the ε4 variant does not mean you have or will develop late-onset Alzheimer's disease. Because of the sensitivity of the result, customers must opt in separately before it is shown.

These ranges are approximate and are drawn from studies of people of European ancestry. The ε4 effect is weaker in people of African ancestry and stronger in people of East Asian ancestry. A 2024 study in Nature Medicine argued that carrying two copies of ε4 should be regarded as a distinct genetic form of Alzheimer's disease, because nearly all such carriers show the biology of the disease by their mid-sixties even when they have no symptoms.

Why APOE is the "main indicator", and why it is not enough

• APOE ε4 is the strongest common genetic risk factor for late-onset Alzheimer's, explaining more risk than the other 75 or so known risk genes combined. That is why it is the one gene a consumer test reports.
• It is a risk factor, not a diagnosis. Many ε4 carriers never develop dementia, and roughly half of people with Alzheimer's carry no ε4 at all.
• The polygenic signal from the rest of the genome adds only a modest improvement, which is why no polygenic Alzheimer's score has FDA authorisation. The evidence is summarised on the genetics page.
• Blood tests for p-tau217, ordered by a clinician, now say far more about whether the disease process is under way than any genetic result.
• No professional body recommends APOE testing for healthy adults outside research, because knowing the result does not yet change treatment. 23andMe's own materials say the same.

Raw data and third-party tools

Customers can download their raw genotype file. It contains the two APOE SNPs (rs429358 and rs7412) and roughly 75 of the rare mutations in APP, PSEN1 and PSEN2 that cause early-onset familial Alzheimer's, but 23andMe labels raw data as not validated for clinical use. Third-party interpreters such as Promethease produce long lists of associations from the SNPedia wiki, many from small or unreplicated studies, and have been known to miscall APOE from some files. Treat anything beyond the official report as research-grade, and confirm any result that matters with a clinical laboratory.

If you have a result and are worried

1. Remember that the result describes risk for a disease of old age; it says nothing about the next year or the next decade.
2. Talk to a genetic counsellor or memory clinic. The one trial of APOE disclosure that showed no lasting harm did so with counselling.
3. Know that the Genetic Information Nondiscrimination Act protects you against health insurers and most employers, but not against life, disability or long-term care insurers.
4. Act on the factors you can change. Blood pressure, hearing, activity and sleep matter at least as much for ε4 carriers as for anyone else.
5. Consider research. Prevention trials actively seek ε4 carriers; see the trials page.